HOMELESSNESS IN THE UNITED STATES, MENTAL HEALTH, ILLICIT DRUG USE, AND CRIME RATES



Ideas:


- Access to mental healthcare contingent on homeless being arrested for crimes

- Homelessness as a result of deinstitutionalization

- Substance abuse/drug arrests are huge part of criminal justice system

- Substance abuse/drug arrests are huge part of criminal justice system

- Facilities for substance abuse treatment

CODE AND ANALYSIS




File contains example code from class with a few annotations.

Codebook for General Social Survey Data

Or if you wish to save the codebook to your computer...




null hypothesis: there is no relation between gender favorability of poor people
my hypothesis: females favor poor people more than males

  • ttest Welch Two Sample t-test data: nes$poorppl by nes$gender t = -4.8338, df = 1898.308, p-value = 1.447e-06 alternative hypothesis: true difference in means is not equal to 0 95 percent confidence interval: -5.950301 -2.515468 sample estimates:
  • mean in group 1. Male respondent selected 72.14640
  • mean in group 2. Female respondent selected 76.37928

  • p-value is 1.447e-06, which is less than .05, so reject the null hypothesis. there *is* a relation between the two variables, it's not likely to get so close to zero, that there is no relation between the two variables.
  • females are slightly more likely to favor poor people than males do.

class(x$attrally) # find what kind of variable this is

summary(x$attrally) # summarizes variable

x$attrally<-factor(x$attrally, levels=c("have not done it and would never do it", "have not done it but might do it", "have done it in the more distant past","have done it in the past yr")) # Re-orders, strings together the categories

x$attrally<-factor(x$attrally, labels=c("1", "2", "3","4"))

x$rallynumeric<-as.numeric(x$attrally) # make your own variable, name it "rallynumeric" because you're changing the factor to a numeric variable

summary(x$rallynumeric)

tab<-table(x$attrally, x$race)
chisq.test(tab)
mosaicplot(tab)
?mosaicplot

install.packages("WDI") #installs World Development Indicators package

library(WDI) #loads the package

WDIsearch("arms") #search for variables involving arms. search any term to find variables involving it.

?WDI #find webpage with code

WDI(country=c("IL","ZW","BT"), indicator=c("SH.XPD.PUBL", "SH.MED.CMHW.P3"), start=1960, end=2011, extra=FALSE) #compare countries based on a particular variable or variables

x$blahblahblah<-x$variablecode #name a variable in the "x" dataframe "blahblahblah"

WDIsearch("health")



MORE INFORMATION AND ARTICLES


Guidelines with R and Research Paper

http://ramnathv.github.com/slidify/

Homeless Information

Arrest Data at US Census Bureau

Substance Abuse Treatment Facilities at US Census Bureau

Healthcare Expenditures at US Census Bureau

Mental Health Data




GROUP MEMBERS:

Name

e-Mail Address

Kirene Holloman
tue68445@temple.edu
Paul Marsh
tue69008@temple.edu
Mike Santucci
tuc09132@temple.edu
Alyssa Dannaker
tue51955@temple.edu


Top